← Field Guide  ·  Tier 2 · Hyperscaler & Vendor References (Cross-Validation)

Salesforce

Enterprise General Intelligence · Agentforce

Vendor · enterprise SaaS4 patterns
In Plain English

This page shows every pattern that Salesforce has published or written about — 4 in total. Each one includes a simple picture of how it works, a short explanation, an example of when to use it, and its trade-offs (the good parts and the not-so-good parts).

Reading patterns from one source at a time matters because it lets you see how a single company or team thinks about building AI agents, and compare their ideas side by side. If a pattern here also appears in the bigger Deep Catalogue, you will see a link so you can read the full, detailed version.

An open reference book with a pencil in the fold under warm lamp light

2C.01Supervisor Agent (hierarchical)

SALESFORCE SUPERVISOR Supervisor Service Agent returns Sales Agent opportunities Case Agent escalations

Salesforce's supervisor pattern: a root agent owns the goal and delegates to specialist agents per topic.

Use caseEnterprise service orchestration across many domain agents.
Trade-offsClear accountability · Supervisor as bottleneck.

See 1B.08

2C.02Skill-Based Routing

SKILL-BASED ROUTING Request Skill Matcher Skill A Agent Skill B Agent Fallback

Agentforce routes requests based on declared skill metadata — agents advertise capabilities; router picks the best match.

Use caseLarge agent portfolios where no single agent fits all cases.
Trade-offsFlexible matching · Skill metadata must be maintained.

2C.03Topic Classifier Agent

TOPIC CLASSIFIER Customer Query Topic Agent Billing Topic Product Topic Escalation

First-stage lightweight agent classifies the user topic and dispatches to the topic-specific agent.

Use caseCustomer-facing deployments with many distinct topics.
Trade-offsSpecialised quality per topic · Classifier errors cascade.

See 1A.03, 1B.07

2C.04Agent-as-Tool

AGENT-AS-TOOL Caller Agent Tool Interface Specialist Agent Return

Specialised agents exposed behind a tool interface so callers compose them without knowing they are agents.

Use caseModular multi-agent systems; legacy agents in new architectures.
Trade-offsClean composition · Performance overhead of extra agent hops.

Cross-ref: Deep Catalogue A2A-ACT-01

Also in Tier 2
Microsoft 7AWS 4OpenAI 3IBM & Others 2